Summary Research Plan of the Overall Program We are currently engaged in a four-year program f research that focuses on the Fries hypothesis, i.e., whether genetic factors limit human life expectancy to 85 years or so. We request five years of continued support for broader research on oldest-old mortality. The overarching theme of the proposed program of research is "Trajectories of Mortality at Advanced Ages". We will address four main issues concerning trajectories of mortality at advanced ages: (i) the environmental and genetic plasticity of mortality rates, (ii) the deceleration of mortality with age, (iii) the interrelationship between disability (ADL's) and mortality, and (iv) male-female differences. To do so: (1) Four the benefit of other researchers we will gather, code, verify, archive, and release public-use files of data on mortality rates and various covariates (such as causes of death and ADL's) for various populations, including males and females in the United States and 27 other countries in recent decades, MZ twins and both same-sex and different-sex DZ twins born in Denmark between 1870 and 1930, and three related fruit-fly species as well as inbred lines of Drosophila. (2) We will use demographic and statistical methods (including life- table, extinct-cohort, smoothing, and survival-analysis methods) to summarize the data and describe trajectories of mortality at advanced ages, including analysis of how mortality rates change with age, over time, for successive birth cohorts, across different countries, across different areas of the United States, for males vs. females, for different causes of death (for Danish twins), at different ADL levels, for different species, for different genotypes, and for population (of Drosophila) that share different chromosomal regions (quantitative trait loci). (3) We will develop alternative theory-based models that incorporate biological knowledge and fit them to the data to gain a understanding of aging and mortality. The models will include genetic and environmental factors, mortality trajectories for individuals, and mortality selection in heterogeneous populations. (4) We will develop demographic and statistical methods of analysis useful for describing the data and analyzing underlying mechanisms as well as biological methods of experimental demography and quantitative trait locus analysis. In particular, we will develop extinct-cohort methods; Lexis-map methods; experimental demography methods for studying ADL's and mortality under different conditions, adaptations of methods of quantitative trait locus analysis for classifying individual Drosophila according to chromosomal regions, models; to simultaneously fit mortality data for the oldest-old in various cohorts, countries, areas of the United States, and species; models to simultaneously analyze mortality data on the general populations of Denmark as well as the MZ and DZ twin population; multivariate stochastic process models to combine mortality data with ADL data, both for Danish twins and for the Medflies and Drosophila, and GoM models of the dynamics of death and disability in the insect experiments.